Methods of forecasting enrollment rate in clinical trial

Abstract

In one embodiment, the present invention provides a method and system of designing a clinical trial enrollment plan, comprising the use of non-linear regression analysis to model the relationship between a pair of clinical trial parameter, e.g., the relationship between N and GSER (Gross Site Enrollment Rate) and the relationship between N and CTER (clinical trial enrollment rate). The values of the other parameter of the pair can be extrapolated from said regression analysis, wherein said extrapolated values of the parameters are outputted as a design or plan product which will be used in clinical trial for improving the performance

What is claimed is: 
     
         1 . A computerized system for determining feasibility of a clinical trial design proposal with proposed values of number of sites (N), number of patients and enrollment time, said system comprising:
 (a) a storage unit for storing a database of clinical trial parameters and their values derived from a plurality of historical clinical trials associated with a disease, wherein said clinical trial parameters comprise number of investigator sites (N) and gross site enrollment rate (GSER), wherein GSER is defined as number of patients enrolled at one investigator site in a unit of time;   (b) a grouping unit for binning the data in said database into a set of bins according to one or more intervals, and tabulating said grouped data by using the median value of N and median value of GSER in each bin;   (c) a computing unit implemented with a regression analysis model for
 (1) conducting a nonlinear regression analysis following said regression analysis model, wherein said regression analysis model fits said tabulated data into GSER=α·e bN +c to obtain a quantitative relationship between N and GSER, wherein a, b and c are constants; and 
 (2) determining confidence level of said clinical trial design proposal with reference to said quantitative relationship obtained from the above (1), wherein said confidence level is determined by using GSER derived from said proposed values of number of sites (N), number of patients and enrollment time for said clinical trial design proposal; and 
   (d) an outputting unit for outputting a result indicating whether said design proposal with the proposed values is feasible, wherein said design proposal is feasible if said confidence level is no less than a desired value.   
     
     
         2 . The system of  claim 1 , wherein said historical clinical trials are similar to said clinical trial in said design proposal when one or more of the following are satisfied:
 a) They are directed to the same or similar disease;   b) They possess similar operational complexity;   c) They have similar inclusion/exclusion criteria for enrolling patients;   d) They have similar number of sites; and   e) They are similar number of patients.   
     
     
         3 . The system of  claim 1 , wherein said desired value is selected from the group consisting of 70%, 80%, 90%, 95%, 98%, and 99%. 
     
     
         4 . The system of  claim 1 , wherein said disease is selected from the group consisting of metabolic disease, respiratory disease, neurologic disease, and cancer. 
     
     
         5 . The system of  claim 1 , wherein the quantitative relationship between N and GSER for clinical trials associated with a single metabolic disease is GSER=1.10·e −3.0193N +0.311. 
     
     
         6 . The system of  claim 1 , wherein the quantitative relationship between N and GSER for clinical trials associated with a single respiratory disease is GSER=0.715·e −3.00533N +0.291. 
     
     
         7 . The system of  claim 1 , wherein the quantitative relationship between N and GSER for clinical trials associated with a single neurologic disease is GSER=0.330·e −0.00482N +0.264. 
     
     
         8 . A computerized system for determining feasibility of a clinical trial design proposal with proposed values of number of sites (N), number of patients and enrollment time, said system comprising:
 (a) a storage unit for storing a database of clinical trial parameters and their values derived from a plurality of historical clinical trials associated with a disease, wherein said clinical trial parameters comprise number of investigator sites (N) and clinical trial enrollment rate (CTER), wherein CTER is defined as number of patients enrolled per site per unit of time;   (b) a grouping unit for binning the data in said database into a set of bins according to one or more intervals, and tabulating said grouped data by using the median value of N and median value of CTER in each bin;   (c) a computing unit implemented with a regression analysis model for
 (1) conducting a nonlinear regression analysis following said regression analysis model, wherein said regression analysis model fits said tabulated data into CTER=A·(1−e BN )+C to obtain a quantitative relationship between N and CTER, wherein A, B, and C are constants; and 
 (2) determining confidence level of said clinical trial design proposal with reference to said quantitative relationship obtained from the above (1), wherein said confidence level is determined by using CTER derived from said proposed values of number of sites (N), number of patients and enrollment time for said clinical trial design proposal; and 
   (d) an outputting unit for outputting a result indicating whether said clinical trial design proposal with the proposed values is feasible, wherein if said confidence level is no less than a desired value, said clinical trial design proposal is feasible.   
     
     
         9 . The system of  claim 8 , wherein said historical clinical trials are those similar to clinical trial in said design proposal. 
     
     
         10 . The system of  claim 9 , wherein said historical clinical trials are similar to said clinical trial in said design proposal when one or more of the following are satisfied:
 a) They are directed to the same or similar disease;   b) They possess similar operational complexity;   c) They have similar inclusion/exclusion criteria for enrolling patients;   d) They have similar number of sites; and   e) They are similar number of patients.   
     
     
         11 . The system of  claim 8 , wherein said desired value is selected from the group consisting of 70%, 80%, 90%, 95%, 98%, and 99%. 
     
     
         12 . The system of  claim 8 , wherein said disease is selected from the group consisting of metabolic disease, respiratory disease, neurologic disease, and cancer. 
     
     
         13 . The system of  claim 8 , wherein the quantitative relationship between N and CTER for clinical trials associated with a single metabolic disease is CTER=363.7·(1·e −0.00124N )+5.58. 
     
     
         14 . The system of  claim 8 , wherein the quantitative relationship between N and CTER for clinical trials associated with a single respiratory disease is CTER=204.6·(1·e −0.00371N )+3.1. 
     
     
         15 . The system of  claim 8 , wherein the quantitative relationship between N and CTER for clinical trials associated with a single neurologic disease is CTER=37.4·(1·e −0.0132N ). 
     
     
         16 . A computerized method for determining feasibility of a clinical trial design proposal with proposed values of number of sites (N), number of patients and enrollment time, said method comprising:
 (a) obtaining values of N and GSER from historical data in historical clinical trials associated with a disease and storing as data into a database of a computerized system, wherein GSER is defined as number of patients enrolled per site per unit of time;   (b) grouping, on a grouping unit of said computerized system, said data into a set of bins according to one or more intervals, and tabulating said grouped data by using the median value of N and median value of GSER in each bin;   (c) conducting, on a computing unit of said computerized system, a nonlinear regression analysis following a regression analysis model implemented to said computing unit, wherein said regression analysis model fits said tabulated data into GSER=α·e bN  c to obtain a quantitative relationship between N and GSER, wherein a, b and c are constants;   (d) determining, on said computing unit, confidence level of said clinical trial design proposal with reference to said quantitative relationship obtained from the above (c), wherein said confidence level is determined by using GSER derived from said proposed values of number of sites (N), number of patients and enrollment time for said clinical trial design proposal, and   (e) outputting, via an outputting unit of said computerized system, a result indicating whether said clinical trial design proposal is feasible, wherein if said confidence level is no less than a desired value, said clinical trial design proposal is feasible.   
     
     
         17 . The method of  claim 16 , wherein said historical clinical trials are those similar to clinical trial in said design proposal. 
     
     
         18 . The method of  claim 17 , wherein said historical clinical trials are similar to said clinical trial in said design proposal when one or more of the following are satisfied:
 a) They are directed to the same or similar disease;   b) They possess similar operational complexity;   c) They have similar inclusion/exclusion criteria for enrolling patients;   d) They have similar number of sites; and   e) They are similar number of patients.   
     
     
         19 . The method of  claim 16 , wherein said desired value is selected from the group consisting of 70%, 80%, 90%, 95%, 98%, and 99%. 
     
     
         20 . A method of selecting one combination of N, number of patients and enrollment time for a clinical trial proposal design from multiple proposals that are feasible under  claim 16 , said method comprising:
 a) Obtaining combinations corresponding to multiple proposals that are feasible under  claim 16 ;   b) Comparing confidence levels of said combinations; and   c) Outputting one combination corresponding to the highest confidence level, wherein said combination comprises N, number of patients and enrollment time.